how do judges decide international relocation cases? · (relocation: shared care arrangement).3 the...
TRANSCRIPT
How Do Judges Decide International Relocation Cases?
Rob George*
Abstract: There has been considerable controversy over the approach taken to international
relocation disputes by the courts of England and Wales, but little has been known about how
first instance judges approach such decisions in ‘everyday’ relocation cases. This article
reports the findings of an empirical study in 2012 which looks at case outcomes and judicial
reasoning in international relocation disputes based on the decisions of trial judges and the
reports of lawyers. The article analyses case characteristics and some factors which may
influence case outcomes, and explores the judicial reasoning used to reach those conclusions.
INTRODUCTION
For over a decade, international relocation disputes have remained one of English family
law’s more controversial issues. These difficult cases involve the application by one parent
for permission to remove their child permanently1 from the United Kingdom against the other
parent’s wishes. Until recently, the leading authority remained the 2001 Court of Appeal
decision in Payne v Payne,2 interpreted in the light of the 2011 Court of Appeal case of K v K
(Relocation: Shared Care Arrangement).3 The most recent decision, Re F (International
Relocation Cases),4 reduces the significance of Payne and calls for a holistic, case-specific
analysis.
While relocation law has been subjected to significant criticism,5 little is known about the
reality of these cases as they are being decided on the ground. The focus of scholars’
attention has understandably been on the principles as set out by appellate judges, together
with the decisions of those few reported first instance cases. Work which has asked about
everyday decision-making in relocation cases has approached the question indirectly, talking
* UCL Faculty of Laws.
This work was conducted when I was a British Academy Postdoctoral Fellow at University College, Oxford,
and I express my thanks to the Academy and to Oxford for their support. I also record my indebtedness to Sir
Mathew Thorpe, who was pivotal in making it possible to collect the data used in this article, along with his
then-Legal Secretaries Victoria Miller and Edward Bennett who did a lot of work to chase judgments for me.
Two sets of reviewers have provided detailed comments on this article, as have the Editors, and it has improved
considerably as a result.
1 Temporary relocation is an increasingly important area, but the issues are quite different from their
permanent relocation counterparts. 2 [2001] EWCA Civ 166, [2001] 1 FLR 1052 (hereafter, ‘Payne’).
3 [2011] EWCA Civ 793, [2012] 2 FLR 880 (hereafter, ‘K v K’).
4 [2015] EWCA Civ 882 (hereafter, ‘Re F (International Relocation Cases)’).
5 See, eg, M. Hayes, ‘Relocation Cases: Is The Court of Appeal Applying the Correct Principles?’ [2006]
CFLQ 351; J. Herring and R. Taylor, ‘Relocating Relocation’ [2006] CFLQ 517; F. Judd QC and R.
George, ‘International Relocation: Do We Stand Alone?’ [2010] Family Law 63; S. Gilmore, ‘The Payne
Saga: Precedent and Family Law Cases’ [2011] Family Law 970; R. George, ‘Reviewing Relocation?’
[2012] CFLQ 110; R. George, Relocation Disputes: Law and Practice in England and New Zealand
(Oxford, Hart Publishing, 2014) (hereafter, ‘Relocation Disputes’).
to practitioners about their experiences.6 This work is valuable, not least in understanding the
wider processes involved in family dispute resolution of which, of course, the family court is
only part. However, such research inevitably gives a somewhat detached perspective, asking
judges and practitioners in the abstract rather than examining actual decisions.
At the same time, a major debate in England and Wales about relocation law is whether leave
to remove a child from the jurisdiction is granted too easily. This important debate has been
significantly hampered by a lack of knowledge about the proportion of applications that is, in
fact, allowed. More generally, little is known about the people involved in relocation
disputes, such as where they are seeking to go and why, how old their children are, and so on.
This paper reports the findings of empirical work carried out in 2012 designed to shed light
on these issues. In particular, the following questions are addressed:
(1) What are the characteristics of international relocation cases decided by courts in
England and Wales, in terms of the families involved and case details?
(2) What are the outcomes of litigated cases in terms of applications being allowed or
refused, and what variables appear to influence those outcomes?
(3) How are judgments constructed in international relocation cases to allow judges to
reach these conclusions?
INTERNATIONAL RELOCATION: THE LEGAL FRAMEWORK
As with all private law disputes about children’s upbringing, the welfare of the child is the
court’s paramount consideration in a relocation dispute,7 supplemented by the welfare
checklist in s 1(3) of the Children Act 1989. In addition, when the cases being discussed in
this article were decided, the courts remained focused on the guidance from Payne, which
highlighted factors of particular relevance to the relocation context. In Thorpe LJ’s judgment,
a so-called ‘discipline’ was set out:8 the discipline asked about the applicant’s motivation and
the level of planning that had gone into the application, followed by the respondent’s
motivations, and then the effect of a refusal of the application on the applicant, all of which
was said to be relevant to an overall analysis of welfare. Similar guidance was given by the
then-President of the Family Division, Dame Elizabeth Butler-Sloss,9 though in the years
following Payne it was Thorpe LJ’s judgment which dominated judicial thinking.
The Payne guidance, particularly Thorpe LJ’s ‘discipline’, was once ‘commonly seen as
walking and talking like a presumption’ in favour of relocation applications being allowed.10
6 Relocation Disputes.
7 Children Act 1989, s 1.
8 Payne, [40]–[41].
9 Payne, [85]-[86].
10 C. Geekie QC, ‘Relocation and Shared Residence: One Route or Two?’ [2008] Family Law 446, 151–52;
see also M. Hayes, ‘Relocation Cases: Is the Court of Appeal Applying the Correct Principles?’ [2006]
CFLQ 351; Relocation Disputes, pp 88-90 and 90-112.
However, more recent decisions – especially K v K (Relocation: Shared Care
Arrangement),11
which was decided shortly before the cases reported in this research were
decided, and Re F (Relocation),12
which was decided three quarters of the way through the
sampling period for this research – made clear that Payne was intended to guide judges in the
factors which they might need to consider, but not as to the outcomes they should reach.13
In
practical terms, this was generally thought to mean that the ‘rigour’ with which judges are
approaching applications increased: the impression that a relocation application is almost
bound to succeed was dying away in 2012 when this research was conducted.14
Consequently, although the guidance itself did not change, it was thought that the way in
which judges approached cases using the guidance did.
The approach to international relocation cases has subsequently shifted again, with the 2015
Court of Appeal decision in Re F (International Relocation Cases).15
Here, the appeal court
said for the first time that following Payne’s guidance may amount to an error of law if the
facts of the case being decided call for a broader (or simply different) approach; the count
commended a global, holistic analysis of all factors, and a parallel analysis of the various
proposals being put forward in an assessment of welfare. While this approach draws heavily
on K v K, it marks a change of emphasis from the position as understood before mid-2015;16
when thinking about the cases discussed in this article, it is therefore important to keep in
mind the legal landscape at the time, and how it appears to have shifted subsequently.
SUMMARY OF METHODOOGY
This paper reports findings from two related studies.17
The first study, which is the principal
source of data, comprises 96 judicial decisions in international relocation cases in the
calendar year 2012, and referred to as the Court Case (‘CC’) sample. The CC sample
comprises 96 international relocation decisions where either the court’s order (N=43), the
court’s judgment (N=53), or both (N=20) were provided. For obvious reasons, judgments
provide a great deal more information than orders alone, though most orders submitted
contained detailed recitals and other information, and even the most basic usually contained
considerable information. Nonetheless, it is worth noting that in the cases where only the
order was available, less information was available and so the analysis that can be performed
11
[2011] EWCA Civ 793, [2012] 2 FLR 880. 12
[2012] EWCA Civ 1364, [2013] 1 FLR 645. 13
See, eg, Re H (Leave to Remove) [2010] EWCA Civ 915, [2010] 2 FLR 1875; J v S (Leave to Remove)
[2010] EWHC 2098 (Fam), [2011] 1 FLR 1694; Re W (Relocation: Removal Outside Jurisdiction)
[2011] EWCA Civ 345, [2011] 2 FLR 409. 14
Relocation Disputes, p 109. 15
[2015] EWCA Civ 882. 16
E Devereux and R George, ‘“Alas Poor Payne, I knew him…” An Interpretation of the Court of Appeal’s
Decision in Re F (International Relocation Cases) [2015] EWCA Civ 882’ [2015] Family Law, October .
For a summary of the law at the time of this research, see R George, F Judd QC, D Garrido and A
Worwood, Relocation: A Practical Guide, 1st Edn (Jordans, 2013), Ch 2. 17
A full methodological appendix, highlighting various limitations of the dataset, is included at the end of
this paper.
on those decisions is more limited. All 96 cases contribute to the quantitative part of this
study, and the 53 cases where judgments were provided underpin the qualitative part.
The second study, designed to augment and test the reliability of the CC sample, comes from
52 questionnaire responses from family lawyers in England and Wales between October 2012
and January 2013, and referred to as the Research Questionnaire (‘RQ’) sample. These 52
cases augment the quantitative part of this paper, but not the qualitative.
Because the CC and RQ samples ask about cases decided in the same period, there is obvious
scope for overlap. Cases were analysed for overlaps, looking at key characteristics such as
number and age of children and proposed destination country. Nine likely overlap cases were
identified in this way.18
When the two samples are analysed separately, all cases are kept in
each sample; however, in places the two samples are combined and then the nine overlap
cases are included only once. Where this is done, the CC cases are used, since the CC sample
is more reliable: the data come directly from court documents, whereas the RQ sample relies
on practitioners’ recollection, potentially some months after their involvement with the case
ended.
QUANTITATIVE FINDINGS
Case Outcomes
England has traditionally been seen as a ‘pro-relocation’ jurisdiction, with applications to
remove a child from the jurisdiction perceived to be treated comparatively favourably.19
However, while some other countries have researched case outcomes to test this question, no
work has previously been done in England and Wales, so making it difficult to know whether
the perception is reflected in practice. The position in Canada (perceived as ‘neutral’ when it
comes to relocation applications) and New Zealand (perceived as ‘anti-relocation’) make for
useful broad-brush comparators, though the different methodologies used for collecting cases
in different studies mean that caution is needed in making direct comparison. In Canada,
Rollie Thompson’s research (which involved only reported cases, though a high proportion
are reported in Canada) showed an overall success rate of 68% for international applications
going outside North America.20
Similarly, Mark Henaghan’s New Zealand study (which
involved all first instance decisions in that jurisdiction) also showed an overall success rate of
68% for international cases in the courts in that country.21
Given that these jurisdictions are
18
A summary of all cases in the two samples, which indicates the likely overlap cases, is contained in R.
George, Relocation Disputes in England and Wales: First Findings from the 2012 Study (Oxford Legal
Research Paper 91/2013), online at ssrn.com/abstract=2306097, Appendix 2 (hereafter, ‘First Findings’) 19
T. Foley, International Child Relocation: Varying Approaches Among Member States to the 1980 Hague
Convention (London, Court of Appeal, 2006); Relocation Disputes, pp 10-12. 20
R. Thompson ‘Heading for the Light: International Relocation from Canada’ (2011) 30 Canadian Family
Law Quarterly 1. Many moves to North America from Canada involve very short distances just across
the border, which may be why differentiating broadly between North American and other destinations is
helpful. 21
M. Henaghan, ‘The Conundrums of Relocation’ (London Metropolitan University Conference, July 2013).
said to represent different approaches to relocation, it is noteworthy that the reported success
rates of applications appear so similar.
The two datasets in this study yield slightly different findings from each other in terms of
overall outcomes: the CC sample had an overall success rate of 71.9%, the RQ cases 59.6%;
when the two datasets are combined (and the supposed duplicates discounted), the overall
success rate is 66.7%. Whichever percentage is used, the comparison with the results of the
(broadly comparable) studies in other jurisdictions, as shown in Table 1, is interesting:
apparently challenging the standard depictions of these jurisdictions as respectively pro-,
neutral, and anti- regarding relocation, the findings show little difference in these broad terms
between the court outcomes in the three jurisdictions.22
Table 1: Table showing the percentage of international relocation applications that
were allowed in the CC and RQ datasets, and in studies from Canada and New
Zealand
Data Set N Percentage of Cases where
Relocation Is Allowed
INT CC
96 71.9
INT RQ, Judicially Determined Cases Only
52 59.6
INT CC and RQ Combined
139 66.7
International Relocation Cases in Canada,
2005-10 (‘neutral’) 47 68.0
International Relocation Cases in New
Zealand, 2011-12 (‘anti-relocation’) 44 68.0
Of course, the results of court litigation do not tell the whole story, because cases which reach
the courts are filtered in various ways by pre-litigation advice and decision-making, all of
which (it is generally assumed) takes place in the shadow of the law,23
influenced by people’s
perceptions of the law as well as non-legal factors and individual bargaining chips. In the
relocation context, earlier research has suggested that the advice that lawyers give to potential
applicant and respondent parents varies considerably from country to country,24
and this is
likely to affect the overall picture of relocation case outcomes. On the other hand, the
proportion of international relocation cases which settle is thought to be relatively low.25
22
None of the differences is statistically significant, whichever of the CC, RQ or combined results is used
for comparison with the Canadian and New Zealand data.
23 R. Mnookin and L. Kornhauser, ‘Bargaining in the Shadow of the Law: The Case of Divorce’ (1979) 88
Yale Law Journal 950 24
Relocation Disputes, esp pp. 85-86. 25
Ibid, p. 3, reporting the views of lawyers in England and New Zealand, both of whom said that it was
unusual to settle relocation cases.
While it is interesting and important to know the overall success rates for international
relocation cases in this sample, the next question is to look at the factors which affect those
case outcomes. It is not the case that every relocation application has an equal chance of
succeeding. Practitioners have long argued that the motivation of the applicant probably is,
and certainly should be, a factor influencing whether the relocation should be allowed or
not;26
and more recently, the English courts have placed renewed focus on the importance of
the child’s existing care arrangements before the relocation dispute begins as a factor which
may influence the decision about relocation.27
Consequently, the analysis now moves to look
at factors which, from a statistical perspective, appear to be relevant to the decision-making
which leads to the overall outcomes reported above.
Factors Influencing Case Outcomes
The study approaches this analysis using different methodological techniques. The first,
called an Information Gain Score (IGS) analysis, ranks each of the variables according to
how much additional information that variable provides when predicting outcomes. The
results of that analysis are reported in an earlier paper,28
their main function being to guide
further work using other methods. The findings of that further work are set out here, divided
into two broad sections.
Families and biographical characteristics
First, we can note that, in line with anecdotal reports, the vast majority of applicants in this
study were mothers. In the CC sample, 93% of applicants were mothers, 5% were fathers,
and 2% were grandparents; in the RQ sample, 95% were mothers and 5% fathers. This
pattern is broadly similar to that in other countries; for example, a Canadian study of
relocation decisions from 2001 to 2010 showed 92% of applicants were mothers, 7% were
fathers and 1% were non-parents.29
One of the most complete pieces of data available in the study is the children’s ages. In the
CC sample (N=140 children), the mean age was 6.75 years, while in the RQ sample (N=88
children) it was 7.9 years. The spread of ages is illustrated in Figure 1, where the percentage
of children of each age in each sample is shown on a bar chart (with percentages used for
ease of comparison). It can be seen that while the RQ sample has a single peak at 8 years, the
CC sample has a higher percentage of children of pre-school age. It is not known why the two
samples would be different in this respect.30
26
See, for example, the argument of Eleanor Hamilton QC, as she then was, in Re B (Leave to Remove:
Impact of Refusal) [2004] EWCA Civ 956, [2005] 2 FLR 239; see also Relocation Disputes, Ch 4. 27
See especially K v K. 28
‘First Findings’, paras 4.9-4.18. 29
N Bala and A Wheeler, ‘Canadian Relocation Cases: Heading Towards Guidelines’ [2012] Canadian
Family Law Quarterly 271, 289. 30
In ‘First Findings’, paras 3.27-3.35, this issue is explored in more detail.
It is also notable that in both samples, children over the age of about 10 or 11 feature
relatively infrequently. When we look later at the qualitative analysis of court judgments, one
reason why this might be the case is that older children’s views are more influential in
decision-making, and so parents may settle litigation (or not start it at all) in some cases
involving an older child, once the child’s views are known.
However, that possible pre-litigation filter issue aside, the data do not support the suggestion
that the children’s ages might affect case outcomes. Indeed, case outcomes are stable
whatever the age of the child involved.31
Figure 1: Bar chart showing the percentage of children of each age in each of the CC
and RQ samples
Another issue to consider is the care arrangements in place for children prior to the relocation
application. One might expect to find that the more the child’s day-to-day care was shared by
the respondent parent, the less likely it would be that a relocation application would succeed.
The intellectual justification for this approach would be that such a child has ‘more to lose’
by relocating, in terms of his or her relationship with the respondent parent, but also that an
existing shared care arrangement would allow the child more readily to remain in the present
location in the main care of the respondent parent if the applicant parent decided to relocate
31
‘First Findings’, paras 4.51-4.55.
0
2
4
6
8
10
12
14
16
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rcen
tage
of
Child
ren
Age of Child
CC Cases (N=140) RQ Cases (N=88)
anyway.32
Table 2 shows the success rates for international relocation applications in the two data
samples.33
Cases were classified as ‘shared care’ where the child spent at least 35% of time
with each parent, ‘overnight stays’ where there was overnight time but less than 35%, ‘direct
contact’ where there was no (or no regular) overnight time, and ‘no direct contact’ where the
child had no face-to-face time with a parent; there was also an ‘other’ category, for cases
where (for example) a non-parent was also involved in the child’s main care, or where the
respondent parent was the main carer.34
In the CC sample, this information was assessed by
the researchers either directly or indirectly from the court documents; in the RQ sample,
respondents reported the care arrangements that had been in place. It can be seen that in the
CC sample, the success rates of applications follow the approximate pattern that would be
expected, with shared care cases least likely be allowed and no contact cases most likely to
be. However, the RQ sample shows almost no difference between the main categories of care
arrangement, which may lead to more doubt about the overall picture. In any case, the
differences are not statistically significant in either dataset, so the differences seen may
simply be the result of random chance.
32
An alternative way of thinking about this issue – in favour of relocation – is to say that a child in a shared
care arrangement has a stronger existing bond with the respondent parent, and so will be better placed to
maintain a positive and beneficial relationship after relocation. 33
Information about care arrangements is not available for all cases in either sample. 34
This classification was not used consistently by judges in their judgments, with the term ‘shared care’ used in
many cases involving overnight time with both parents regardless of the extent of sharing.
Table 2: Table showing success rates for international relocation applications in the
CC and RQ samples, and in the two combined, for different pre-relocation child care
arrangements
Care
arrangement
CC Cases RQ Cases CC and RQ Combined
N Success
rate (%) N
Success
rate (%) N
Success
rate (%)
All cases
where care
arrangements
known
71 66.2 51 60.8 118 61.9
Shared care
arrangement
(35% <)
8 50.0 16 56.3 24 54.2
Some
overnight
stays
43 60.5 22 50.0 63 55.6
Direct contact,
no overnight
stays 11 72.7 6 50.0 16 62.5
No direct
contact 5 100.0 3 100.0 8 100.0
Other care
arrangements 4 100.0 4 100.0 7 100.0
That caveat – given the inevitable limitations of the dataset – aside, these findings might
incline us to suggest that cases involving no overnight contact with the respondent parent are
more likely to be allowed than those involving at least some overnight stays. However, it is
more difficult to know whether, once that line has been crossed, the amount of sharing of
care makes any difference: the CC data suggest that it does, but the RQ and combined data
are more equivocal. Further work is needed to clarify this issue. Interestingly, as noted below
in the court judgments analysis,35
judges seldom varied their language in describing the
relationship between the child and the non-moving parent. Save where that parent presented
some kind of risk to the child, they were commonly described as, for example, having ‘a very
loving and close relationship’, whatever type of care arrangement was in place.
A separate issue which the data seem to suggest may be of some relevance is the applicant
parent’s current relationship status, whether married, cohabiting, and so on.36
As the figures
in Table 3 show, it seems that in this sample those applicants who were in some kind of
stable relationship had a higher chance of their applications succeeding than those who were
35
Below, text after fn 81. 36
There is no apparent variation in relation to respondent parents’ relationship status, as might be expected:
‘First Findings’, para 4.57.
not in an on-going relationship.37
Table 3: Table showing success rates for international relocation applications in the CC
sample for different current relationship statuses of the applicant parent
Current relationship type
APPLICANT PARENTS
N Success rate (%)
All cases where relationship status
known 47 68.1
Married
17 82.4
Engaged
2 [100.0]
Cohabiting
4 [75.0]
Non-cohabiting relationship 4 [75.0]
No on-going relationship 20 50.0
[ ] = small numbers, therefore percentages should be interpreted with caution
Looking in more detail at the applicants who are married or engaged compared with those
who are not in an on-going relationship, some striking differences emerge. For example,
amongst the 19 single applicants from whom this information is known, none was a British
national seeking to emigrate, whereas 14 of the 19 married or engaged applicants were.
Similarly, the vast majority of single applicants were seeking to relocate in order to return
home (80%) or to get increased family support (80%); between them, these reasons
accounted for nine out of ten cases involving single applicants. Of the remaining two cases,
one was based on a job offer abroad,38
while the other was seen as a pure ‘lifestyle’
decision.39
Amongst the married applicants, on the other hand, the clear majority were bringing an
application based primarily on a factor related to their new spouse. In some cases, he was
already in the new location (N=6), while in other cases he was seeking to return to his
original home country (N=5). Four of these eleven cases also involved the new partner
37
The result for those applicants not in an on-going relationship is statistically significant at the 5%
confidence interval, with a p-value of 0.022. 38
Case CC-81. The mother was offered a new job with her existing employer in its Brussels office, having
moved to London from another EU country in 2007. 39
Case CC-75. Note, though, that a background of domestic violence and child abuse by the father made this
case atypical as ‘lifestyle’ applications go.
having a job offer, and job offers for the partner were the main basis for a further three cases.
A few applications amongst the married group were based on the applicant herself seeking to
return to her original home country (N=3) or were pure ‘lifestyle’ cases (N=2).
It is interesting to wonder how best to understand these patterns. Are judges concerned to
protect applicants’ new relationships, which some authorities suggest will struggle if a new
partner is prevented from pursuing his life ambitions involving relocation, so as to try to
avoid the children experiencing a second relationship breakdown?40
Do some judges take the
view that children need a father-figure in their lives, and so keep children of single applicants
near the respondent fathers while allowing re-partnered applicants to move because this role
is seen as being performed by the step-father?41
Alternatively, is it that judges are generally
less impressed by applications premised on a desire by the applicant to return ‘home’, but
find those cases motivated by new partners or their careers more compelling?
Curiously, this aspect does not feature strongly in judicial reasoning. Judges frequently talked
about the effect of a refusal of the application, but rarely linked this discussion to issues
relating to new partners. As we will see,42
the focus was more often on the applicant’s present
life in this jurisdiction and, in many cases, a need for support in the proposed destination
country, neither of which tended to be linked to new partners.
Relocation Destinations and Motivations
The second broad category of findings relates to the destinations of the proposed relocations,
and the parties’ motivations in seeking or opposing such a move. Starting with proposed
relocation destinations, most international applications involve to one of three groups of
countries: countries within the EU, countries in North America, and countries in Australasia;
for the purposes of this analysis, other destinations are treated together as a collective ‘other’
group. Table 4 shows the success rates for applications in these four groups for the CC and
RQ data sets, and for the two combined.
40
See, for example, Re B (Removal from Jurisdiction); Re S (Removal from Jurisdiction) [2003] EWCA Civ
1149, [2003] 2 FLR 1043, [11]-[12] and [31]-[32]. 41
See generally R. van Krieken, ‘The “Best Interests of the Child” on Parental Separation: On the
“Civilizing of Parents”’ (2005) 68 Modern Law Review 25, 30-31 and 35. 42
Below, text after fn Error! Bookmark not defined..
Table 4: Table showing success rates for relocation applications in the CC and RQ
samples, and in the two combined, for different proposed destinations
Proposed
relocation
destination
CC Cases RQ Cases CC and RQ Combined
N Success
rate (%) N
Success
rate (%) N
Success
rate (%)
All cases
where
destination
known
95 72.6 51 60.8 139 67.6
European
Union 35 80.0 20 50.0 54 68.5
North
America 21 71.4 13 61.5 31 64.5
Australia /
New
Zealand
23 52.2 10 [60.0] 32 56.3
Other
destinations 16 87.5 8 [87.5] 22 86.4
[ ] = small values, therefore percentages should be interpreted with caution.
The CC data, on the left side, appear to tell a fairly coherent story: put simply, the further
away the proposed destination, the less likely it was that the relocation would be allowed.
The RQ cases, on the other hand, tell a different and in many ways less clear story. Proposed
moves to the EU here had the lowest success rate (50.0%), and there is little difference
between North American cases (61.5%) and Australasian cases (60.0%). The big variation in
outcomes for EU cases between the CC and RQ data is particularly puzzling, since this group
is the largest in both datasets and ought, therefore, to be least vulnerable to random
fluctuations. Looking deeper, though, there may be reason to think that the RQ sample of EU
moves is populated with ‘unusual’ cases. For example, while just 5.7% of the CC sample of
EU moves (N=35) were heard in the High Court (either by full time Judges or by Deputies),
almost a third (31.3%) of RQ cases to the EU were heard in that court (N=16). Consequently,
there may be reasons to think that the CC sample is more ‘typical’,43
and so to give greater
weight to the higher success rate for EU applications seen in that data. This interpretation also
fits with a more natural explanation of the patterns in the data, since it makes more sense if
comparatively short-distance moves within the EU should be allowed more frequently than
proposed relocations to the other side of the world.44
Turning to applicants’ reasons for seeking to relocate, earlier research in other countries has
43
Few cases are heard by High Court Judges, and almost all of them involve moves outside the EU unless
they have unusual complexities. For a third of all EU cases to be heard by High Court Judges suggests a
sample of cases which were unusual or complex on their facts, possibly caused by research participants
more readily remembering an ‘interesting’ High Court case than a ‘normal’ case before a lower level of
judge. 44
Conversely, language and cultural differences may be less with a long-distance Commonwealth country than
some short-distance European ones where the applicant and children are originally British nationals – but that
may be less of a factor where applicants want to return to their original home country.
shown that most applicants give several reasons.45
Consequently, while we can look, for
example, at all cases where the applicant was seeking to ‘return home’ to her original
country, and at cases which involved a move for a new job, some cases will fall into both
groups.
Table 5 breaks down cases by reference to the three main reasons given: returning home,
moving to take up a new job that the applicant or her new partner had been offered, or
moving for a generally improved lifestyle; plus a catch-all ‘other reasons’ category. Since the
categories are not mutually exclusive, the total number of cases for each dataset is less than
the sum of each reason for moving.
Table 5: Table showing success rates for international relocation applications in the
CC and RQ samples, and in the two combined, for different primarily reasons given by
the applicant for seeking to relocate
Reason for
seeking
relocation
CC Cases RQ Cases CC and RQ Combined
N Success
rate (%) N
Success
rate (%) N
Success
rate (%)
All cases
with known
reasons
75 72.0 52 59.6 123 65.9
Return to
applicant’s
original
country
42 73.8 23 73.9 61 72.1
New job for
applicant or
partner
16 93.8 25 56.0 40 70.0
Lifestyle
choice 21 47.6 21 52.4 40 47.5
Other
reasons 9 [66.7] 5 [40.0] 14 57.1
[ ] = small values, therefore percentages should be interpreted with caution
NB: Categories of reasons for seeking relocation are not mutually exclusive, therefore the sum of all reasons is greater than
the total number of cases.
These data again suggest a somewhat confused picture between the two datasets, with the CC
data suggesting that moving for a new job has a significantly higher chance of success while
moving to go home is about average, whereas the RQ data show the exact opposite. Both
datasets show that lifestyle choice cases are less likely to succeed than any other reason,
though, and this finding is statistically significant in both the court case and combined
datasets. This finding is consistent with practitioners’ views from earlier research suggesting
45
See N. Taylor, M. Gollop and M. Henaghan, Relocation Following Parental Separation: The Welfare and
Best Interests of Children (Dunedin, Centre for Research on Children and Families, 2010), p. 26; R.
Kaspiew, J. Behrens and B. Smyth, ‘Relocation Disputes in Separated Families Prior to the 2006
Reforms: An Empirical Study’ (2011) 86 Family Matters 72, 74; P. Parkinson, J. Cashmore and J.
Single, ‘Mothers Wishing to Relocate with Children: Actual and Perceived Reasons’ (2011) 27 Canadian
Family Law Journal 11.
that ‘lifestyle choice cases are the hardest to [succeed in bringing]’.46
Finally, we can look briefly at the way in which some of these factors appear to interact.47
Most interesting is the interaction between the proposed destination and the applicant’s
principal reason for seeking to relocate, as set out in Table 6.
Table 6: Table showing success rates for relocation applications in the CC sample,
divided into the main geographic locations of the proposed destinations, for different
primarily reasons given by the applicant for seeking to relocate
Reason for
seeking
relocation
European Union
North America Australia / New
Zealand
N= Success
rate (%) N=
Success
rate (%) N=
Success
rate (%)
All cases
where reasons
known
28 78.6 16 68.8 18 50.0
Return to
applicant’s
original home
country
18 83.3 10 70.0 7 [42.9]
New job for
applicant or
partner
5 [100.0] 2 [100.0] 5 [80.0]
Lifestyle
choice 6 [33.3] 5 [80.0] 10 40.0
[ ] = small values, therefore percentages should be interpreted with caution. NB: Categories of reasons for
seeking relocation are not mutually exclusive, therefore the sum of all reasons will be greater than the total
number of cases.
The pattern is not entirely consistent between the three groups of destinations, and the small
numbers in many of the categories mean that significant caution is needed when looking at
many of the percentages. But the patterns amongst these cases are largely unsurprising, given
the previous data seen. For example, permission to move to Australia in ‘lifestyle’ cases was
particularly unusual, whereas ‘going home’ to EU countries was very often allowed.
QUALITATIVE ANALYSIS OF COURT JUDGMENTS
So far, the analysis in this article has focused primarily on individual variables and how they
may impact on case outcomes. In practice, of course, each case involves many variables
which interact. As noted at the end of the previous section, some multi-variable analysis can
be done in a statistical way, but that analysis can inevitably only go so far, because the
46
Relocation Disputes, p. 102. 47
It is not possible to do this for all variables, as the data are not clear enough to lead to robust conclusions.
outcomes of court cases depend upon considerations which simply cannot be quantified. If
we knew that factors x, y and z added up led to outcome 1, and factors a, b and c added up led
to outcome 2, there would be little need for courts at all. The reality is that many of the
considerations which inform judicial reasoning and hence court outcomes are not quantifiable
in this way. In particular, the personalities of the people involved often have an effect – even
a decisive effect – on the judge’s decision.48
Consequently, we turn now to look at judges’
reasoning as expressed in judgments as a way better to understand how various factors fit
together in leading to case outcomes.
In looking at the 53 judgments available for this analysis, a number of themes can be
identified. First, it is interesting to see how the judges constructed the law itself, particularly
as this issue is particularly difficult and, perhaps, controversial following K v K.49
From there,
we turn to look at particular issues raised by the Payne guidance which are of especial
relevance to the overall reasoning processes seen in the judgments.
Constructions of the law
The treatment of the Payne guidance
In general, judges did not engage in lengthy discussion of the law relevant to relocation cases.
Most cases quoted the key summaries from Payne,50
and many judges proceeded to address
each of the points raised using headings based on either Thorpe LJ’s ‘discipline’ or Butler-
Sloss P’s ‘guidance’ (or sometimes both). Other authorities were mentioned intermittently.51
Most judges also gave express consideration to the items on the welfare checklist.52
While judges frequently reiterated that ‘[i]t is imperative ... that the court heeds the words of
caution of the Court of Appeal in K v K and avoids the risk of elevating the guidance in
Payne to the status of principle or, worse still, presumption’,53
only occasionally did they
make evaluative comments about the authorities. One District Judge discussed at some length
various criticisms of Payne before somewhat begrudgingly saying that ‘I shall ... submit to
the “discipline” suggested by Lord Justice Thorpe’.54
The only point on which there was any real variation in approaches was about cases
48
Relocation Disputes, p 173. 49
See, eg, E. Devereux and R. George, ‘When Will the Supreme Court Put Us Out of Our Payne?’ [2014]
Family Law 1586. 50
Payne v Payne, [40]-[41] and [85]-[86]. 51
The most commonly cited was K v K, followed by J v S (Leave to Remove) [2010] EWHC 2098 (Fam),
[2011] 1 FLR 1694 and Re AR (A Child: Relocation) [2010] EWHC 1346 (Fam), [2010] 2 FLR 1577; in
the final months of 2012, judges also sometimes cited Re F (Relocation) [2012] EWCA Civ 1364, [2013]
1 FLR 645, which was decided in October 2012. 52
Children Act 1989, s 1(3). 53
CC-34, [86]. 54
CC-46, [27]; also CC-58, [7].
involving shared care arrangements, which was described by one judge as a ‘particularly
vexed question’.55
There were two clear views in the eight cases where this issue was
discussed.56
One was that the court’s approach would be different in a shared care
application, and therefore that ‘[w]hether or not this is a shared care case is a question, in my
judgment, which I need to deal with at the outset of any discussion of the relevant law’.57
For
these judges, a finding of shared care meant that the Payne guidance was often given little
further attention, whereas if the judge declined to label the arrangements as such then Payne
was the court’s focus. It is interesting to wonder whether some judges might have used this
labelling as a means of reinforcing a decision already made on another basis: for example, if
a judge were inclined to refuse a relocation application for other reasons then categorising the
case as ‘shared care’ might help to bolster that conclusion if Payne is seen as not applying to
shared care cases. There is no way to assess whether this is so from court judgments, but
earlier work suggests that judges may be conscious of these considerations.58
Other judges rejected the idea that shared care cases called for a different legal approach –
‘the court should not categorise cases in accordance with the concepts of primary or shared
care’.59
In a clear reflection of the arguments put forward by counsel, judges in this group
often commented that they were ‘not assisted ... by attempting to decide who was the primary
carer’.60
These judges avoided any classification, but were happy to state (as Black LJ had
done in K v K) that ‘the guidance set out in Payne carries less weight’ in cases involving truly
shared care.61
It is likely that this confusion over shared care relocation cases has been resolved by the
October 2012 decision in Re F (Relocation),62
wherein Munby LJ reiterated the non-
classificatory approach and declared that Moore-Bick and Black LJJ represented the majority
in K v K,63
and more so by Ryder LJ’s decision in Re F (International Relocation Cases).
However, it is clear that throughout 2012 there was variation in approaches to these cases,
and judges felt compelled to engage with these questions.
Human rights law
55
CC-81, [104]. 56
Not all of the cases discussing this issue were classified as shared care cases – judges sometimes discussed
the controversy but then determined that the facts of the case fell outside the shared care area.
Consequently, the 8 cases here do not correspond exactly with the 8 shared care cases shown in Table 2;
in 3 of those cases, only an order was sent to the study, so no comment can be made about the judges’
approaches in their judgments. 57
CC-23, [42]; also CC-27, [7]; CC-12, [33] and CC-54, [29]. 58
Relocation Disputes, pp 69-71 and 78-79. 59
CC-96, [5]. 60
CC-61, [36]. 61
CC-67, [19]. 62
[2012] EWCA Civ 1364, [2013] 1 FLR 645 63
On the confusion over the ratio of K v K, see R. George, ‘Reviewing Relocation?’ [2012] CFLQ 110.
Another legal point, conspicuous primarily by its absence, was discussion of human rights
issues. Very few cases even mentioned the rights of the children, parents or others involved;
those that did contained only perfunctory comments. The handful of cases where rights were
mentioned fell into two groups. In one group, the mention of rights involved a purely abstract
acknowledgement that the parties had rights which were engaged by the decision:
I am also acutely conscious of each of the parties’ Article 6 and Article 8 rights under the
human Rights Act and, in coming to my decision, I have all of that at the forefront of my
mind.64
The other group of cases involved judges going one step further and noting (still only in the
abstract) that any interference with Article 8 rights needs to be necessary and proportionate
and in the pursuit of a legitimate aim. Two cases – which mark the high point of the rights
analysis in the relocation judgments – put the issue in these ways:
This order does affect each party’s rights to family life under the European Convention.
The interference in the father’s is merited by the welfare interests of his son and is
proportionate.65
[In allowing relocation] I would interfere with the father’s right to family life pursuant to
Article 8 ECHR, which I can only do if I consider it necessary and proportionate. I do
consider it necessary and proportionate to do so.66
It would be easy to criticise these judgments for their minimal engagement with the
potentially substantial interferences with the rights of the parties involved. But these judges at
least commented expressly, albeit minimally, on the human rights issues in their judgments.
In practice in the relocation cases in this study, therefore, human rights issues were not a
central focus for judges. However, human rights issues have a potentially substantial effect
on analysis of the issues.67
It is not that judges should undertake a rights analysis as a sort of
pro forma, but their analysis of ‘the entire family situation’ needs to demonstrate that they
have engaged in an overarching balancing of interests.68
A mere acknowledgement of rights
or assertion of proportionality may not advance matters much, and may not be adequate to
allow an appeal court to review the judge’s ‘compliance or otherwise’ with human rights
obligations.69
That said, although the Court of Appeal in Re Y suggested that no separate
analysis of Article 8 is required in private law children cases,70
in Re F (International
Relocation Cases) Ryder LJ held that in a case involving a potential international relocation
64
CC-55, [79]; similarly CC-27, [A31]; CC-41, [5]; CC-94, [5]. 65
CC-62, [34]. 66
CC-69, [98]. 67
J. Herring and R. Taylor, ‘Relocation Relocation’ [2006] CFLQ 517. 68
YC v United Kingdom (Application No 4547/10) [2012] 2 FLR 332 (ECtHR). 69
Re B (Care Order: Proportionality: Criterion for Review) [2013] UKSC 33, [2013] 2 FLR 1075, [45]. 70
Re Y (Removal from Jurisdiction: Failure to Consider Family Segmentation) [2014] EWCA Civ 1287,
[2014] 3 FCR 483.
the potential for an interference with Article 8 rights means that a clear and separate
proportionality evaluation may be required. Quite what this involves is not spelt out by Ryder
LJ, but experience in the public law field suggests that a careful assessment involving the
parallel analysis of all realistic options can be done in a way which involves the
proportionality assessment and overall balancing exercise required under Article 8.71
Factors coming from Payne
The applicant’s proposals
Both judgments in Payne ask about the applicant’s proposals. Judges usually scrutinised the
substance of applicants’ relocation proposals in great detail. For example, in many cases
where relocation was refused, judges made critical remarks about the details of the plans.
While occasionally these criticisms were substantive,72
a more common criticism was that the
plans were ‘vague’,73
‘ill thought out and ... over-ambitious’,74
or ‘not well evidenced’.75
While, of course, these remarks may have been entirely justified, a plan that one person sees
as ‘over-ambitious’ could be considered exciting and potentially highly rewarding by another.
For example, in another case the judge said of the mother:
I accept that she may not have dotted every ‘i’ and crossed very ‘t’, but frankly ... the
mother seems to me to be so capable that any unexpected obstacle could – and would – be
easily overcome by her.76
In the end, while judges spoke at greater length about practicalities, their views about the
each parent’s truthfulness and reliability often seemed more influential.
The broad pattern was that unsuccessful parents were the subject of criticism from judges
about what they had claimed or the way in which they had claimed it. For example, in about
half the cases where relocation was refused, judges were concerned about the applicant
parent’s attitude to contact and her trustworthiness in relation to the child’s on-going
relationship with the father. One or two cases appeared to turn entirely on these findings,
such as CC-55, an application to move with children aged 8 and 7 to the United States. The
judge found that the ‘cunning and manipulative’ mother ‘seeks to undermine contact’; her
‘very abusive’ comments about the father were ‘indicative of the depth of hatred’ she felt.77
71
The European Court of Human Rights sometimes takes the same view: see, eg, YC v United Kingdom
(Application No 4547/10) [2012] 2 FLR 332 (ECtHR). 72
For example, CC-13 involved the mother proposing to leave her job in London and go to Canada, where
she had no other connections and where she would commute for nearly 3 hours each day for 4 hours of
low-paid work which would barely cover the costs of childcare. 73
CC-90, judge’s cover note. 74
CC-55, [87]. 75
CC-39, [5]. 76
CC-67, [20]. 77
CC-55, [77], [51], and [70] respectively.
On the other hand, cases where relocation was allowed were characterised by adverse
remarks about the respondent parent’s attitude to the child or to the applicant parent,78
often
coupled with negative comments about his evidence at the hearing. The latter of these ranged
from mild remarks about the father ‘not being completely frank’,79
through more significant
findings that the father was ‘a dishonest man’,80
and giving evidence which was ‘in parts
excessive and in parts false’.81
The effect of relocation on the child’s relationship with the non-moving parent
Payne next asks about the effect of a relocation on the non-moving parent, and about the
steps that can be taken to offset the loss of relationship with the child, about which judges
expressed a variety of views. With the exception of those handful of cases where the fathers
represented a safety risk to the children, judgments almost invariably spoke of there being ‘a
very loving and close relationship’,82
where ‘it goes without saying that this child has a strong
bond with her father’.83
Indeed, so similar were the comments across the body of cases that,
looking just at these sections of the judgments, it would be difficult to differentiate between
those applications that were granted and those that were refused. A similar convergence of
language is seen in those cases where judges spoke about the loss of the ‘everyday parenting
experience’ – what Eleanor King J once described as ‘lying around on a sofa on a Saturday
evening, eating pizza and watching a DVD with your dad, or being taken to school by him
every other Monday’.84
By contrast, a small number of cases proceeded on a basis which seemed to lack much
appreciation of the importance of these ordinary everyday interactions. Case CC-44 was a
particularly egregious example, where the child had been mainly looked after by the father on
a day-to-day basis for a year. Allowing the mother to remove the child from the father’s care
to join her abroad, the judge said this:
Furthermore, and I know that the father will find it difficult to accept this, it seems to me
that [the child] will have a better relationship with his father in many ways if regular
contact takes place. I did get the impression ... that an awful lot of time that [the child]
spent with his father was either at school or being fed or doing homework and not much
of it was time for [the child] and his father to have fun together ... in what one might call
78
For example, a drug-dealer father who had been in prison for most of the child’s life claimed ‘that he had
been a good role model to [his son, which] demonstrated ... a lack of understanding of the impact of his
lifestyle on his then-family’: CC-62, [10]. 79
CC-81, [31]. 80
CC-5, [28]. 81
CC-12, [29]. 82
CC-72, [10]. 83
CC-7, [82]. 84
J v S (Leave to Remove) [2010] EWHC 2098 (Fam), [2011] 1 FLR 1694, [99].
down time. This will be available to his father on contact.85
Many would find this a truly extraordinary statement. Of course a large part of the week for a
school-aged child involves being at school, having dinner and doing homework. The father
was continuing to work full-time in his job, since that was the family’s only financial support,
and weekends were often being spent having contact with the mother. To tell the father that
losing the day-to-day care of his son would lead to ‘a better relationship’ was, at best,
unfortunate language.86
Here we see direct engagement by the judge with the father, but the
sexist undertone and detachment from the father’s own experiences as claimed in evidence
are likely to have undermined the intended positive effect.
Much the same sentiment can be found in case CC-56, though the case did not involve the
same direct engagement between judge and parent. Here, the judge accepted that the father
had been providing 65% of the children’s care for the three years since the parents’
separation. He had given up full-time work to do so, while the mother continued in her
demanding job in London and started a relationship with a man (to whom she was now
married) who lived in the USA. As the judge said, ‘[m]ainly her career but also her current
marital choice have impacted upon her ability physically to spend time with [the children]
though she has done her utmost’.87
Somehow, the fact that ‘a very high proportion of [the
mother’s] contacts have been at extended weekends and thus there was proportionately more
time available with the children’88
counted for more than the father’s daily commitment,
because the mother was said to be providing a ‘pivotal emotional relationship’ such that ‘the
welfare of the children is threatened by [her] inability to cope and thereafter to provide
emotionally’.89
Despite the judge’s recognition that ‘[the father] undoubtedly loves his
children and has been a good father to them, shouldering as he has the greater bulk of the
parenting duties’,90
still the inherent qualities of the mother – who might alternatively have
been thought to have behaved selfishly and without thought for the children – made her more
suitable to care for them and justified a removal from the father’s care and from the
jurisdiction.
Where the judge allowed relocation, there was usually a statement to the effect that this
father-child bond was strong enough to survive the resulting geographic separation. As one
judge put it, ‘the fundamental strength of the relationship ... would not be impaired’.91
In
another case, the judge thought that ‘[t]he quantity of contact will be reduced but the quality
85
CC-40, [26]. 86
See also CC-48, [111]: ‘I am sure [the children] will miss [the father] and he will miss them, but this will
make their regular reunions all the more satisfying for them.’ 87
CC-56, [24(ii)(b)]. 88
Ibid, [23]. 89
Ibid, [47]. 90
Ibid, [53]. 91
CC-48, [111].
will not in my judgment be affected’.92
Judges had ‘no doubt that the father will be
devastated if permission to relocate is given’93
and endeavoured to make orders for on-going
contact that reflected the fact that ‘[the] father remains hugely important ... and he will see
[the children] just as often and for as much time as is possible’.94
By contrast, a number of judgments in which relocation was refused expressed concern that
the relocation might risk a loss of the meaningful relationship between the child and the
father. In some cases, these concerns were based on a lack of confidence in the mother’s
commitment to contact, but more often the perceived risk was based on a view that the
distance involved and the consequent change in the nature of the relationship would be too
serious for a meaningful relationship to survive.
A majority of cases where this loss or damage was seen to be a ‘decisive point’ in refusing
relocation involved an existing strong relationship between child and father,95
but in others it
was thought that the child’s relationship with the father needed strengthening, and that
relocation would prevent this from taking place; here, the relocation was prevented largely in
order to improve a currently poor relationship.96
These two classes of comment might suggest certain judicial (or legal) assumptions about the
benefits of a strong father-child relationship.97
Relocation might be refused because of an
existing strong relationship which was already important to the child, or might be refused
because there was no such relationship and it needed to be built and strengthened.
Conversely, relocation might be allowed because the parent-child bond was strong and would
allow the relationship to survive the distance, or it might be allowed because the bond was
not that strong and so the risk of loss was less important because the relationship was not as
central to the child’s welfare as other factors.
These judicial choices explicitly relate to the law’s present uncertainty about parenting norms
and the importance of fathers in children’s upbringing. In 2014, after the cases in this study
were decided, the Children Act 1989 was amended to include a so-called ‘parental
involvement presumption’.98
This provision states, in short, that a parent should be involved
in his or her child’s life unless it is not safe for that to happen, but that ‘involvement’ does not
mean any particular division of time. While this new subsection has not yet been the subject
of significant judicial consideration, either in general or in the relocation context specifically,
Ryder LJ has said that in an international relocation case s 1(2A) is likely to ‘heighten the
92
CC-12, [35]. 93
CC-96, [14]. 94
CC-61, [39]. 95
For example, CC-30 and CC-84. 96
CC-36 and CC-55. 97
See, eg, R. Bailey-Harris, J. Barron and J. Pearce, ‘From Utility to Rights? The Presumption of Contact in
Practice’ [1999] IJLPF 111. 98
Children Act 1989, s 1(2A), inserted by the Children and Families Act 2014.
court’s scrutiny of the arrangements that are proposed by each parent’.99
Effect of refusal on the applicant
Set against this, judges considered the likely impact that a refusal of the application might
have, where again judges seemed to be able to construct identical arguments leading to
opposite conclusions. The issue of the effect of refusal is, perhaps, the area which most tests
trial judges in the relocation context. In the wake of several Court of Appeal judgments in the
early 2000s,100
a narrative was established that judges were required to find that there would
be a substantial impact of refusal and that this factor would carry substantial weight in the
overall analysis.101
While recent decisions have stressed that this is not the case,102
in an area
like relocation, where most judges are inexperienced and may see a case only once every two
or three years, it takes time for these changes to filter down. Consequently, judges may still
see it as ‘safer’ to allocate considerable weight to the impact of refusal, while a finding that
there would be little or no impact – or, even more so, that there would be impact but that it is
outweighed in the overall evaluation – remains ‘braver’.103
It is therefore noteworthy to find many judges pointing out – in line with the more recent
authority104
– that the effect on the parent of refusing relocation was only one of many factors
to be assessed based on the evidence in the case:
I give no primacy to the mother’s application and do not adopt any approach which
creates some presumption in her favour by attaching weight to the emotional and
psychological impact upon the mother, or any consequential impact upon the child. These
factors do, nevertheless, have to be considered as part of the overall balance.105
In assessing a possible shift away from the one-time focus on the effect of refusal of leave,
two trends can be noted. One, in the cases where relocation was refused, involves judges
willing to find either that the impact of refusal would not be as bad as the mother thought, or
that she was able to deal with that effect well enough.
[A]lthough the mother would be very disappointed ... she would cope and ... she would
not allow her disappointment to taint the loving relationship that either parent has with
99
Re F (International Relocation Cases), [35]. 100
In particular, Re B (Removal from Jurisdiction); Re S (Removal from Jurisdiction) [2003] EWCA Civ
1149, [2003] 2 FLR 1043 and Re B (Leave to Remove: Impact of Refusal) [2004] EWCA Civ 956, [2005]
2 FLR 239. 101
M. Hayes, ‘Relocation Cases: Is the Court of Appeal Applying the Correct Principles?’ [2006] CFLQ
351, 366. 102
Re W (Relocation: Removal Outside Jurisdiction) [2011] EWCA Civ 345, [2011] 2 FLR 409; K v K. 103
Relocation Disputes, p 79. 104
At the time, the clearest expositions were J v S (Leave to Remove) [2010] EWHC 2098 (Fam), [2011] 1
FLR 1694 and K v K. The point is reiterated by the general requirement of a holistic, non-presumptive
approach from Re F (International Relocation Cases) [2015] EWCA Civ 882. 105
CC-21, [29].
the child. She would move on. Indeed, she said so herself in oral evidence.106
[T]he mother will undoubtedly be extremely upset... [T]he Cafcass officer was of the
clear view that the mother would cope ... [and] I formed the same view. ... I feel sure that
the mother will cope in this country, as she has done for many years.107
These cases – including some which seemed to involve judges instructing mothers that they
were required to cope108
– illustrate the willingness of judges to find that the overall balance
of the welfare assessment favours refusal of relocation, despite the effect on the applicant.
The competing trend, where relocation was allowed, shows that a successful application can
be brought without any reliance on the effect of refusal. In a small but substantial minority of
cases, ‘the mother’s case is not put on the basis that she will suffer severely if not permitted
to travel’.109
As another judge explained:
This is not the kind of relocation case in which a mother states that she cannot continue
with life in England... When asked how she would feel about the refusal of her
application, the mother in effect said that she would just have to cope with it. ... In fact,
unusually for an applicant, her situation in England is, I believe, considerably tougher
than she admits.110
Set against these cases were a sizeable number of applications which were allowed where
judges clearly considered that the impact of refusal was a major consideration. In many of
these, the judges linked their conclusion to specific parts of the evidence:
If she was not allowed to go the nightmare ... that her life ha[s] become would continue,
and I do not think that is an unfair way of describing it.111
Mother was trying to hide how distressed she feels at her situation here... Her desire to go
home and her devastation if she has to stay here were plainly genuine. ... I do think she
will be devastated by a refusal. She was trying to be brave ... [but] she is trapped and
worn down by life here.112
If refused permission she would be very distressed and frustrated and would entirely
blame the father. ... She would be badly affected if refused permission, with an indirect
effect on [the child].113
106
CC-67, [21]. 107
CC-85, [87(e)]. 108
For example, CC-30, [81]. 109
CC-14, [41(5)]. 110
CC-27, [A51]. 111
CC-5, [40]. 112
CC-49, [40] and [56] respectively. 113
CC-62, [25] and [30].
However, in other cases, the conclusion that the mother would be ‘devastated’ seemed less
compelling. In one case, the parents were operating a shared care arrangement in central
London and the mother’s job at a large multinational company took her to Brussels on a
regular basis; she proposed to move there full time, though her employer was content to have
her based either in Brussels or in London. The judge’s view in this case that the mother
‘would be extremely upset, indeed devastated, if she could not move to Brussels’114
seems,
with respect, to overstate the case. The mother, irritated no doubt, would in all likelihood
have carried on her life much as before, and perhaps in time sought a posting that required
less travel (surely a possibility, given that she worked for a major international company with
its global headquarters in London).
This point was well made by another judge, who observed that ‘hyperbole has become
jargon’115
at the start of eight full paragraphs dealing with the ‘devastation’ issue:
Both counsel, and I am bound to say I, sighed when we heard the word ‘devastated’ yet
again. The problem is that there is a limited vocabulary of words available to describe the
extent of a person’s feelings of despair, loss, grief, fear, exasperation, frustration and
disappointment, but in so saying I illustrate that there are other words than ‘devastated’.
In the mother’s case, she told the father that the refusal of her wish to relocate would be a
disaster. She would lose her identify, autonomy and independence; and, inevitably, she
would be devastated. I found that to be a considerable overstatement of what is likely to
be the truth.116
That is not to say that the judge discounted the effect on her – it was considered and weighed,
and the application was allowed. This judge was not alone, however, in expressing concern at
the dominance of claims of ‘devastation’. Another judge, who also allowed the relocation
application, found expressly that ‘M[other] will feel distress – but not devastation, as she said
in evidence – if she does not [relocate]’.117
A number of cases where relocation was
ultimately refused also commented on this point, with the word ‘disappointed’ featuring often
as a likely impact; in one particularly telling case the maternal grandmother, giving evidence,
‘thought her daughter would be devastated if leave were refused, but conceded that the word
“devastated” was suggested to her by the mother’s lawyer’.118
Children’s wishes and feelings
An issue not mentioned expressly in the Payne guidance, though imported into it by its
114
CC-81, [79]. 115
CC-48, [112]. 116
Ibid. 117
CC-72, [17]. See also CC-67, [21]: ‘the mother would be disappointed with refusal, but I hesitate to use
the word “devastated” which was employed during the hearing as this seems to me to be putting it far too
high’; CC-73, [65]: ‘I accept that the mother would be upset and distressed ... but ... I do not accept that
she would be devastated’. 118
CC-36, [28].
reference to the welfare checklist, is the views, wishes and feelings of the children involved.
Such wishes and feelings are principally brought into the family justice system through an
officer from the Child and Family Court Advisory and Support Service (Cafcass) or an
Independent Social Worker’s report. Such reports can be limited to the issue of the child’s
wishes and feelings, though they tend to be much more wide-ranging and encompass all
aspects of the case. A full welfare report typically carries great weight in parenting disputes;
if the report includes a recommendation, judges are required to give specific reasons for
rejecting that view if they reach a different conclusion.119
However, it has long been thought
that Cafcass reports are less influential in relocation cases than elsewhere in child law.120
Indeed, in the cases in this study, reports were not invariably ordered, often because the
parties and the judge simply felt they did not need one.
Where Cafcass reports were made, judicial reactions were mixed. One area in which judges
were generally united, though, was that the report was important in relaying the wishes and
feelings of the children involved. However, in practical terms, children’s views had quite
limited impact in the judges’ reasoning in most cases. Since the mean age of children in
relocation cases is just under 7 years, it is perhaps unsurprising that judges often simply noted
that the children were ‘too young for [their] wishes to assist me’,121
or even that they were
‘too young to express [their] wishes and feelings on the issues before the court’.122
For children in middle childhood (roughly ages 5 to 10), the most common approach was to
say that the issues were too complex for their wishes and feelings to carry much weight.
While one may have some sympathy with this view, it should be remembered that the
children were not being asked to decide the case, but rather to express their wishes and
feelings about it. While these children may not ‘realise fully what would be involved’ in
relocation,123
some showed real insight into what was being decided. One 7 year old who had
been living in England for about three years rather wisely said to the Cafcass officer that ‘he
missed his life in [the other country] but that if he was [there] he might just as well have been
saying that he missed his life in England’.124
The judge thought that these expressed views
‘don’t help me very much’ in making the decision, but one might have inferred that the child
was aware of the benefits of his present life and was not expressing any desire to change.125
Indeed, in other cases, children of this age did express views which influenced the decision-
making. Usually these were cases where the children’s views accorded with the outcome that
119
See, for example, Re V (Residence: Review) [1995] 2 FLR 1010 (CA). 120
Relocation Disputes, p. 103. 121
CC-41, [41]. 122
CC-74, [69]; also CC-53, [47]. 123
CC-39, [14]. 124
CC-40, [10]; also CC-72, [25], where another 7 year old had written two letters to the judge which
revealed ‘that [the child] is very torn indeed’. 125
CC-40, [25]; see also CC-68, where the 11 year old child made the seemingly important point that she
had never even visited the proposed destination country and ‘thought she should go on holiday there first
to see if she liked it’ (para [4]).
the judge had reached, such as Case CC-69, where the parents were in a high-conflict shared
care arrangement which the judge thought was not working well. The judge reported this
exchange when the 8 year old met the judge in person:
I asked him ‘is there anything you want to say?’ And he said ‘I want to live with my
mum’ and because he said that I felt I had to ask him ‘why?’ He said ‘when I see my
mum my tummy doesn’t ache any more, and when I see dad my tummy aches’.126
Unsurprisingly, in the handful of cases involving older children judges often said that their
views ‘are something I have to give a lot of weight, albeit they are not determinative’.127
One
13 year old spoke for many children when reporting that ‘she would, herself, [have] preferred
not to have been asked’.128
Despite that, the judge found significant value in what the child
said:
she told [the Independent Social Worker] that she would be fine either way, [and]
expanded that what she really wanted for herself (and I underline this for emphasis) was
to move to the United States and go to school there, but she would miss her dad and
would prefer it if dad moved there too and lived nearby.129
Thus case outcomes generally accorded with older children’s wishes, but there were
exceptions. In one case, relocation was granted in relation to a 13 year old boy who was
described as ‘very apprehensive’ about the proposed relocation and who was ‘close to tears’
when discussing it with the Cafcass officer.130
Two cases involved the judges specifically
discounting the views of older children, one because ‘she tells people what they want to hear
– she is a people-pleaser’,131
and the other because the judge concluded that the children’s
views had been unduly influenced by their mother and were therefore simply expressing her
views rather than their own.132
Overall, it is clear that in only a small minority of cases did the children’s views have any
great influence on the decision. In a high proportion of cases their views were either not
sought or were entirely dismissed – the child was ‘too young’ or their views were discounted
for some other reason – and even in those cases where weight was being given to the child’s
views, the impression from the judgments was that the decision would have been the same
regardless of the child’s view. In other words, judges said that they were giving weight to the
child’s views when those views aligned with their own conclusions about the outcome of the
126
CC-69, [76]. Note that judges are strongly discouraged from obtaining evidence directly from children:
Family Justice Council Guidelines for Judges Meeting Children subject to Family Proceedings 2010; Re
KP (Abduction: Judge Meeting Child: Conduct of Interview) [2014] EWCA Civ 554. 127
CC-64, [17]. 128
CC-48, [125]. 129
Ibid, [124]. 130
CC-33, [6] and [26] respectively. 131
CC-43, [13]. 132
CC-85, [75].
case (a reinforcement approach), and found reasons to dismiss the child’s views when the
outcome was not going to fit with them.
CONCLUSIONS
This is the first study in England and Wales that gives any insight into the outcomes of
international relocation cases in the family court. While the two parts of the study show
slightly varying results in places, the overall picture is clear and informative. We see, most
crucially, that it cannot be said that relocation is automatically allowed – in this sample at
least, something like 30 to 40% of applications were refused. We can also identify some
variables which, overall, may be related to those case outcomes, such as the distance of the
proposed move and the applicant’s current relationship status.
Moreover, we have been able to look in detail at the reasoning that goes into making a
determination in a relocation case. Presented with enormous amounts of evidence on wide-
ranging issues, trial judges have many choices to make. They have to decide what is relevant
at all and in what ways it is relevant; they have to choose between competing (often directly
contradictory) versions of the same events; and they have to make predictions about various
possible futures for a child, with effects potentially lasting a lifetime. They then have to
construct a narrative account of their choices which weaves together their findings and
conclusions within the legal framework so as to allow the decision to be understandable,
coherent and (ideally) with minimal scope for appeal.
The research reported in this article tells us a great deal about the decision-making processes
of judges in everyday relocation cases. We can see, in particular, that judges give very serious
consideration to two main competing factors – the effect on the applicant of refusing her
application, and the effect on the child’s relationship with the respondent if relocation goes
ahead. These are assessed against the background considerations of the practicalities of the
proposal (including considerable focus on the plans for on-going contact), and the
motivations of both parents.
In most cases, judgments were constructed such that all these aspects lined up to point
towards the same conclusion. Sometimes that was, no doubt, an accurate reflection of the
way in which the case had developed, but other cases raised some doubts, given that the same
issue could be constructed to lead to exactly opposite conclusions: for example, a strong
existing relationship between the child and the respondent could be seen as a reason to refuse
a relocation (more to lose) or to allow one (relationship strong enough to survive the move).
By contrast, a handful of judgments expressed genuine anxiety and uncertainty, which may
more accurately reflect the reality of many relocation cases.
More generally, the big question about relocation when these cases were decided was
whether the Court of Appeal’s reinterpretation of Payne in K v K went far enough,133
or
133
This is Mostyn J’s current view, according to NV v OV [2014] EWHC 4130 (Fam), [7].
whether the guidance itself needed to be rewritten. This research suggests a rather mixed
answer to that question.
A number of concerns were expressed about the ‘Payne in the light of K v K’ approach.
Perhaps the main concern was that Payne was premised on there being a primary carer
applicant, when many cases no longer fit that pattern. K v K (as understood after Re F134
)
demands that judges deal with this issue by weighing the Payne considerations differently,
but there was a question as to whether Payne lends itself to such flexible interpretation.135
Here, the findings of this research are somewhat equivocal. Some judges clearly understood
that they were permitted to take this approach and did so with little apparent difficulty. In
other cases, though, it appeared that Payne was applied fairly uncritically, despite being
inappropriate to the facts of the case.136
Whether because of greater seniority, greater
experience, less risk-aversion,137
or some other consideration, some judges readily modified
the Payne guidance while others followed it closely as if it were a manual. This, in short, was
the difficulty: there was a concern that K v K risked an inconsistency of approach and
therefore of like cases not being treated alike. Moreover, as relocation cases are increasingly
allocated to the District Bench, the Court of Appeal’s ability to have an overview of the
general approach diminishes, since most applications for permission to appeal will now be
heard by one of dozens of Circuit Judges rather than the handful of specialists in the Court of
Appeal.
It may be that the Court of Appeal’s decision in Re F (International Relocation Cases) in
August 2015 has changed the picture somewhat. Here, Ryder LJ suggested that undue
reliance on Payne’s guidance could amount to an appealable error of law, and that the
‘required reading’ for a judge in an international relocation case no longer included Payne,
but instead came principally from K v K and Re F (Relocation). Even in the lifetime of the
cases in this study, Re F (Relocation) appeared to be helping to standardise understandings of
the approach after K v K, and it may be that Re F (International Relocation Cases) goes even
further in standardising a holistic, all-factor approach.
Nonetheless, it may be wondered whether Payne will continue to exert influence, since it has
still not been over-ruled. K v K, which Ryder LJ posits as the central case, is itself focused on
the continued relevance of Payne, and thus working back through the authorities one still
ends up with the guidance from that case. The concern, looking at the reasoning of judges in
this study, is not that the issues raised by Payne are inappropriate in many cases; rather, it is
the manner in which those questions are asked and the assumptions which underpin them
(particularly in relation to the child having a single primary carer who is inevitably the
134
Re F (Relocation) [2012] EWCA Civ 1364, [2013] 1 FLR 645. 135
R. George, ‘International Relocation, Care Arrangements and Case Taxonomy’ [2012] Family Law 1478;
E. Devereux and R. George, ‘When Will the Supreme Court Put Us Out of Our Payne?’ [2014] Family
Law 1586. 136
CC-40 and CC-56 were the most blatant examples. 137
On judges’ concerns about appeals, see Relocation Disputes, p 79
mother) which seem inappropriate, and asking judges to determine the extent to which that
guidance is relevant on a case-by-case basis may be too much of a stretch.
Indeed, it is difficult to see why the senior judiciary remains so attached to Payne. It cannot
really be about precedent, because Payne is only guidance and the Court of Appeal has
regularly abandoned one set of guidance in favour of another – that, indeed, is how Payne
became the leading case in the first place. Despite the clarification from Re F (International
Relocation Cases), it is time for the Court of Appeal or the Supreme Court to look at this
issue again from scratch, ideally by taking a number of conjoined appeals, with the aim of
issuing new guidance that applies to all cases without relying on individual judges to make ad
hoc modifications.138
138
My proposed guidelines are set out in Relocation Disputes, pp 157-9.
METHODOLOGICAL APPENDIX
This appendix describes the way in which the dataset was generated, and the resulting
limitations of the data in terms of their representativeness or otherwise of relocation cases as
a whole which reduce the extent to which they may be relied upon to make inferential
conclusions about this class of case generally based on statistical significance of any of the
findings.
The Court Case Sample
The CC sample of cases was decided in the calendar year 2012 and gathered directly from
trial judges and practitioners.139
Judges were contacted directly by email and letter. Initial
contacts were made by Thorpe LJ, with follow-up communication through the Family
Division Liaison Judges and the Designated Family Judges. Individual judges were also
contacted personally where I received a ‘tip off’ that they had heard a relocation case
recently. Individual judges were very helpful in sending materials when asked, but of course
it is not possible to know how many decided a case but did not remember to send it in, or
chose not to do so. Following a low response rate in the early part of the study, lawyers were
then contacted in various ways to ask for their assistance, once it was clear that the
President’s authorisation for the project allowed anyone who had a judgment or order to
submit it to the project.
The CC study was an attempt to gather the entire population of relocation cases heard in
England and Wales that year. It is not suggested that this attempt was successful;140
96
eligible cases were collected,141
but it is difficult to assess what proportion of the total
population of relocation cases this is. Official records give some insight into how many
international relocation cases are heard by the courts each year. Ministry of Justice statistics
show that, in 2012, 384 children were the subject of orders allowing them to be removed
from the UK following private law proceedings.142
However, caution may be needed in
approaching this number. First, it is not clear how the MoJ data are collected, and so the
likely accuracy of the number is difficult to assess. Second, it relates to the number of
children, not the number of cases. There were 1.5 children per case in the CC sample which,
if typical, would suggest something like 260 cases in total. Third, the MoJ data do not appear
to differentiate between permanent and temporary leave to remove a child from the
jurisdiction. Nonetheless, this number gives some insight into the size of the CC sample
139
Sir Nicholas Wall, President of the Family Division when this project was running, authorised access to
these cases. 140
Because both judges and lawyers were being asked for cases, a good many of the 96 were collected more
than once. Many judges also reported, directly or through Thorpe LJ’s office, that they had not heard any
relocation cases in 2012. 141
A number of other cases with international elements were sent in, but discarded as not being relocation
cases within the definition used in this study. 142
See http://www.justice.gov.uk/downloads/statistics/courts-and-sentencing/csq-q4-2012/csq-q4-2012-
family.csv, column DR.
relative to the total number of cases. Assuming the figure of 384 children to be correct, and
assuming 1.5 children per case on average, and assuming that holiday cases might amount to
between 0 and 20% of those cases (though the number is unknown), the 96 cases in the CC
sample would represent between 36 and 46% of cases. It is not possible to know whether the
cases are representative or not, though having both lawyers and judges sending cases goes
some way towards mitigating the risk of intentional bias on the part of those submitting
material, since the incentives for the two professional groups to do so may differ. While a
handful of judges and lawyers sent more than one case, the vast majority of individuals
sending material sent just one case, and some cases were received more than once from
different sources.
The Research Questionnaire Sample
In the final three months of 2012, a research questionnaire was distributed to family law
practitioners throughout England and Wales. The questionnaire was intended to provide an
alternative source of data about cases decided in 2012, to triangulate findings against the CC
sample and check for possible biases in the case sample submitted to the study. It was also
useful for bolstering the overall number of cases being considered, though both direct
collection of cases for the CC sample and the questionnaire study for the RQ sample faced
significant difficulties in obtaining responses from busy professionals.
Most questions asked lawyers to provide detailed information about their most recent
relocation case, designed to elicit information comparable with the data collected from the
court cases themselves. Practitioners were asked to complete a questionnaire about their most
recent relocation case, international or domestic, if they had done one in the previous 3 years
(and also to send nil returns), and the questions then differentiated between type of case, year
of decision, and mode of resolution (court adjudication or settlement). Some aspects of the
RQ data suggest that some respondents may have reported their most recent ‘interesting’ case
(however they saw that), since – for example – the sample appears in places to be skewed
towards High Court decisions.
The questionnaire was prepared in hard copy and online. 1,000 hard copies were distributed
at practitioner conferences and by direct mail to chambers and family law firms around
England and Wales, selected to get a range or large and small firms / chambers in all areas of
the country. The online version was also advertised in articles and news items posted in
practitioner journals and websites, and through the practitioner organisations Resolution, the
FLBA, and the ALC. 187 completed responses were received, including 8 nil returns. While
this number suggests a low response rate, it should be borne in mind that relocation cases are
not everyday work for family lawyers, and a great many practitioners will not have done one
in the last three years, even if they do private law children work. Of the 187 total responses,
52 related to judicially determined international cases heard in 2012, and it is these 52 cases
which are used in this paper, being directly comparable with the 96 CC cases.
Data Analysis
This paper draws on a mixture of quantitative and qualitative methods. For the purpose of the
quantitative analysis, each case was entered into a database containing 84 fields, though no
single case had data relevant to every field. These fields included basic information such as
the number and ages of the children, the proposed destination country and the case outcome,
but also more detailed information about process (such as location of court, type of judge,
whether each party had a lawyer), the family (such nationality of parents, presence of wider
maternal and/or paternal family in this country and in the proposed destination), and the
outcome (such as the extent of contact ordered). Many fields simply did not apply to many
cases (such as whether there were child protection concerns). But particularly where only an
order (i.e. no judgment) was available,143
there were cases where more common fields could
not be completed (such as the parents’ nationalities). Consequently, when it comes to analysis
of the data, it is not always possible to consider all cases in relation to particular variables.
The quantitative data were then analysed by Dr Ornella Cominetti, a statistician then at the
Oxford University Department of Statistics. She used three primary techniques to assess the
data: Pearson’s chi-square test, logistical regression models, and information gain score (IGS)
analysis. While the first two are well known, IGS is a less common technique. IGS evaluates
the worth of each variable by measuring the information gain of a variable to the class
(where, in this research, the outcome variable is whether the relocation is allowed or not).144
The qualitative analysis related to the 53 cases from the CC sample in which court judgments
were available. The judgments were analysed using a broadly grounded methods theory, but
one informed by the fact that many of the key issues being sought were known in advance
based on the research questions – since the questions were partly about the use that is made
of the Payne guidance, issues identified within that guidance formed part of the coding
framework from the start.
143
Some questionnaires also had questions left blank, or respondents had written ‘don’t know’ next to a
question. 144
See M. Hall and G. Holmes, ‘Benchmarking Attribute Selection Techniques for Discrete Class Data
Mining’ (2003) 15 IEEE Transactions on Knowledge and Data Engineering 1437, 1447.